How to get ChatGPT and Perplexity to cite your company (an AEO guide)

More and more clients start the buying process not with Google, but with a question typed into ChatGPT or Perplexity. They get a finished answer with cited sources. Being one of those sources is a new kind of visibility, and you can compete for it on purpose.

That is the job of AEO, Answer Engine Optimization. Below we show what answer engines look for and how to build content they choose as a source.

How AEO differs from SEO

SEO fights for a position on a results list. AEO fights for being the source inside a single answer.

That distinction changes how you write. In SEO, part of what matters is that the user clicks and stays on the page. In AEO, what matters is that the model can lift a specific, complete answer from your text and attribute it to you. The two goals don’t clash, but they need a different content structure.

Rule 1: a direct answer at the top of each section

Answer engines like a paragraph that can be lifted and pasted as a finished answer, without reading the whole text.

In practice, under each heading you place one to three sentences that directly answer the question in that heading. Only then do you expand. If the reader or the model has to dig through five paragraphs to find the answer, the content loses to a competitor who gave it immediately.

Rule 2: headings as questions, not labels

People ask AI in full sentences. Your headings should answer those full questions.

Instead of a heading “Costs” write “How much does an ERP implementation cost”. Instead of “Process” write “What the implementation looks like step by step”. The model matches the user’s question to a heading in the content, so the closer your heading is to a real question, the higher the chance of a citation.

Rule 3: unique knowledge that exists nowhere else

Models filter out content rewritten from other sources. First-hand knowledge wins.

This is the most important and hardest-to-fake rule. Data from your implementations, specific processes, numbers you won’t find in ten other articles, opinions grounded in experience. Given a choice between ten texts repeating the same thing and one with real, unique content, the answer engine points to that one. The same thing your clients are looking for.

Rule 4: structure a machine understands

AI handles a wall of text badly. It handles structure well.

Lists, tables, clear definitions, short paragraphs, a sensible heading hierarchy. Structured data in the page code, such as FAQ or Article schema, gives the model an extra signal about what the content is. The clearer the structure, the easier it is for the model to lift a finished answer.

Rule 5: topical consistency across the domain

One great article isn’t enough. Answer engines trust domains that cover their field systematically.

That is topical authority. If you have ten related articles around one topic, the model treats your domain as a source in that niche, not a random blog. So AEO is not a one-article task, it is a cluster strategy. We expand on this in the article on topical authority.

How to check whether it works

AEO monitoring is simpler than it sounds. You regularly type key industry questions into ChatGPT and Perplexity and check whether and when your domain shows up as a source.

That gives concrete feedback: which topics you already own and which still belong to competitors. You plan the next content on that basis.

Where to start

You don’t need to redo the whole blog at once. Start with five questions your clients ask most often and write content for them following the rules above.

If you want, we’ll do it together. As part of a free one-month content plan we check who AI cites in your industry today and point out the topics that get you into those answers fastest.